THE UNIVERSITY OF CALGARY An Event Structure Analysis of Derived Verbs in Engiish by Naomi Kei Sawai A TEiESIS SCTBMïlTED TO THE FACULT'Y OF GRADUATE STüDlES IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS DEPARTMENT OF LINGUISTICS CALGARY, ALBERTA AUGUST, 1997 O Naomi Kei Sawai 1997 National Library Bibliothèque nationale I*m of Canada du Canada Acquisitions and Acquisitions et Bibliographie Services services bibliographiques 395 Wellington Street 395. nie Wellington OttawaON K1AON4 OttawaON K1AON4 Canada Canada The author has granted a non- L'auteur a accordé une licence non exclusive licence allowing the exclusive permettant à la National Librq of Canada to Bibliothèque nationale du Canada de reproduce, loan, distribute or sell reproduire, prêter, distribuer ou copies of this thesis in microform, vendre des copies de cette thèse sous paper or electronic formats. la forme de microfiche/nlm, de reproduction sur papier ou sur format électronique. The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts kom it Ni la thèse ni des extraits substantiels may be printed or otherwise de celle-ci ne doivent être imprimés reproduced without the author's ou autrement reproduits sans son permission. autorisation. ABSTRACT This thesis presents a case snidy of morphologically cornplex verbs in English formed by -ifyl-ize affxation. In particular, we examine the argument smicture and event structure properties of these verbs and outline a theory of where in the grarmnar these properties are derived. We observe that, while -@/-ise verbs cannot be identified as having consistent argument stnrcture properties, they do have a uniquely identifiable aspectual propeq, which is that they denote delimited events, in the sense of Temy (1994). Hence, we propose an event stnictu~analysis of the verbs in which we analyze the affixes as AFFIXES which serve as morphological markers of delimitation. Following ment proposds conceming the structural representation of events, we suggest that the affixes head an aspecnial projection in the phrasal syntax labelled FP-delimit. Finally, we examine the implications of our proposal for a universal theory of event structure representation. That 1have corne this far in my acadernic weer is the result of an abundance of support, encouragement and positive afhrmations that 1have received £rom rnany people in my He, for whom 1will always be thM. 1 WUbegin with Elizabeth Ritter. Not only has she served as my thesis supervisor, she has served as a mentor. Through her boundiess energy and entiusiasm for the field, her interest in my own work and progress, and her unceasing support (reading through never- ending cirafts of every chapter), she has inspired me to WRllE! and aven me the confidence and the courage to FINISH! 1 will never have enough words to express my deeply-felt gratitude. 1would iike to hedythank the other members of my thesis commitiee -- John Archibald, who encouraged me immensely in the beginning stages of this research (helping me with a pilot test that 1 tried out last year) and Amanda Pounder, who conmbuted to this work by chaüenging me with insightfd questions and painstakingly edi~gthis thesis. Mer people 1 would like to thank are Penelope Codding for taking the heto discuss crystallize with me, Heidi Harley for talking with me about her work (and for writing such enjoyable papers), Vi for keeping me organized and the other members of the department, who have encouraged me over the years. 1 am especially grateful to EihGuilfoyk, who, when 1 was an undergraduate student here, urged me to continue studying and leanwig. To alI rny fellow students - in particular, Leah for king such a fantastic office mate, Susan and Julie for their words of encouragement, Grace for giving us much-needed breaks in her lovely home and Val for her positive energy -- Thanks for all of your support! For financial support, without which this project would not have been possible. 1 have several sources to thank - the Department of Linguistics for the teaching and research assistantships, the Dean's Office of the University of Calgary for the Spatial Masters Scholarship, the Govemrnent of Alberta for the Rovince of Albata Graduate Scholarship, and finally, 1 gratefully acknowledge the support provided by SSHRCC grant #410-94- 0478 CO Elizabeth Ritter. Last, but not least, special thanks go to my husband Wayne for putting up with my wacky behaviour over the past few years and for sustaining me with food while 1 worked in the late hours of the night (or early hours of the morning). To my precious brother Kenji, who lifted my spirits with ail the emaiis from Hawaii. To my dad (the "wise white-haired man") for sharing his humorous views about the English language. To my mom for her constant encouragement and prayers. And to Goci, who listend Dedicated to my Parents, who betieved in me. TABLE OF CONTENTS Ovemiew of mesis ------ ---------------------- 21 1.4.1 Chapter Two: Against the LEXICAL AFFM Hyp~thesis---------------- 21 1-4.2 Chapter Three: The E~ENTAFFIX Hypothesis ---------------------- 21 1.4.3 Chapter Four. The Syntactic Representation of Events----------------- 22 1.4.4 Chapter Five: Extensions of the Event Affix Hypothesis -O---------- 22 Chapter Three The EVENTAFFIX Hypothesis -------------------------------45 Functor matesand Aspecmal Liceising ---------------------- -- 46 3.1.1 A Preliminary Discussion of Argument Licensing ----------------- 47 3.1.2 Non-mematic Arguments ---- 48 3.1.3 -ifLl-ii as Functor---- --------------- ------------------ 50 Chapter Four The Syntactic Representation of Events----------------- 89 4.1 Theoretical BackWund --------------------------------------- ------- 90 4.1.1 L-Syntactic Event Smcture (Hale and Keyser (1993))-------------- 90 4.1.2 me Role of AspP (Borer (1993)) -------------------- ----------- --- 92 4.1.3 Syntactic D-Events (Riüer and Rosen (1997)------------------------ 95 4.1.4 The Organization of Clau* Sm-----------------------------98 ----- ----- ---- ---------- 117 4-4 Zero Derivation --O - ----- --- 4.4.1 The Aspecmai Roperties of Zero-Derived Verbs ---------------- 119 4.4.2 A Syntactic Analy sis of Zero-Derived Verbs ..................... 122 Chapter Five Extensions of the Event Affix Hypothesis------------ 131 CHAPTER ONE Introduction : Deriving the Properties of Morphologically Complex Verbs in English 1.0 PREAMBLE Our domain of inquiry concerns the meaning of derived verbs, specifically those derived by -ifil-ize affixation. The question we seek to answer is 'Where do the (uniquely) verbal properties of &rive. verbs corne fiorn?' It is reasonable to assume that a derived verb inherits at lest some of its semantic content from its nominal or adjectival stem However, these verbs, like non-denved or primitive verbs, have panicuiar semantic properties not associated with nouns and adjectives. Verbs belong to the grammatical category that expresses the action, state or process of a sentence (Genmer (1981), Frawley (1992) and Levin (in prep.)). In other words, verbs t@dy encode events, 'a relatively temporal relation in conceptual space' Frawley (1992:144). Encoded in îhe meaning of verb, therefore, is the notion of rime. Furthemore, the eventuality that the verb denotes invoIves one or more semantic participants (or arguments), entities that either perform or are affecteci by the action, state or process named by the verb. These arguments appear as the grammatical subject and object in the syntax, A noun, on the other hand, refers to an entity it is the thing that participates in the action denoted by the verb, and generally does not require participants of its own. An adjective typically denotes a smte or attribute and as such, does require a semantic argument (it must be attributed of something). However, an adjective does not entaii a temporal relation. Thus, verbs are uniquely disringuished ikom nouns and adjectives in that they entail an eventive (temporal) notion as weii as a number of semantic participants. 2 The defining semantic pmpenies of verbs are perhaps best introduced by examining the mehg of a simple, nonderived verb. Consider the example below illusmting the primitive verb put.' (lb) illustrates that the meaning of the verb is decomposable into more primitive semantic predicates like CAUSE, GO and AT (Jackendoff (1974, 1990), Camer and Randal1 (1993), Levin and Rappaport Hovav (to appear)). Each predicate requires the presence of one semantic participant, which is represented as a variable (either x, y or 2). The variables in tum cmbe defined by the sernantic role they play in the event These semtic roles are iisted in (lc). In ou.exarnple, CAUSE specifies an AGENT argument (x), GO a 'ïNEME argument (y) and AT a GOAL argument (z). The AGENT is the entiry that actively engages in the activity of putting. The THEME is the entity that undergoes the action, and the GOAL is the entity towards which the action is duected. The verb put in the above example, denotes an eventuality in which an AGENT causes an entity, the THEME, to undergo movement such thai the THEME cornes to be at a location or GOAL. Central to this definition then is the specification of the arguments, as well as the sernantic predicates CAUSE, BE and AT, which specify the relations between the arguments. 1 For ease of exposition, in order to illustrate the decomposition of the meaning of put into its component parts, we utilize here a type of predicate decompositional representation of rneaning. This is for explanatory purposes only. It does not mean that we subscribe to a particular theory of argument projection in which such a representation is assumed to be present in the lexicon dong with specified principles of rnapping. 3 Like the primitive verb put, a derived verb also denotes the action of a sentence and thus functions to set up relationships between arguments. Consider the examples listed in (2) and (3) below which illustrate -iO affixation and -ize affixation respectively.